The interaction between social protection and agriculture. A review of evidence

The interaction between social protection and agriculture A review of evidence

The interaction between social protection and agriculture A review of evidence Nyasha Tirivayi, Marco Knowles and Benjamin Davis FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS Rome, 2013

The From Protection to Production (PtoP) project is financed principally by the UK Department for International Development (DFID) and the Food and Agriculture Organization of the UN (FAO), with additional support from the European Union. The PtoP project is part of a larger effort, the Transfer Project, joint with UNICEF, Save the Children and the University of North Carolina, to support the implementation of impact evaluations of cash transfer programmes in sub- Saharan Africa. The designations employed and the presentation of material in this information product do not imply the expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United Nations (FAO) concerning the legal or development status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or products of manufacturers, whether or not these have been patented, does not imply that these have been endorsed or recommended by FAO in preference to others of a similar nature that are not mentioned. The views expressed in this information product are those of the author(s) and do not necessarily reflect the views or policies of FAO. FAO 2013 FAO encourages the use, reproduction and dissemination of material in this information product. Except where otherwise indicated, material may be copied, downloaded and printed for private study, research and teaching purposes, or for use in non-commercial products or services, provided that appropriate acknowledgement of FAO as the source and copyright holder is given and that FAO s endorsement of users views, products or services is not implied in any way. All requests for translation and adaptation rights, and for resale and other commercial use rights should be made via www.fao.org/contact-us/licence-request or addressed to copyright@fao.org. FAO information products are available on the FAO website (www.fao.org/publications) and can be purchased through publications-sales@fao.org.

Acknowledgments This literature review was prepared by Nyasha Tirivayi, Marco Knowles and Benjamin Davis (FAO Agricultural Development Economics Division). The paper benefited greatly from comments received at the FAO Agricultural Development Economics Division seminar and contributions provided by: Elisenda Estruch, Vito Cistulli, Ana Paula de la O Campos, Cristina Rapone, Carol Djeddah and Dubravka Bojic (Gender, Equity and Rural Employment Division); Alessandro Romeo, Federica Damiani, Frédéric Dévé and Shogshagsherry Ajemian, (Agricultural Development Economics Division); Johanna Jelensperger, Joana Borrero and Martina Park (Nutrition Division); Pamerla Pozarny (FAO Investment Centre); Lahsen Ababouch, Rolf Willmann and Nicole Franz (Fisheries and Aquaculture Policy and Economics Division); Daniela Kalikoski, Susana Siar and Francis Chopin, (Fisheries and Aquaculture Resource Use and Conservation Division); and Tomas Lindemann (Climate, Energy and Tenure Division). We are grateful to our principal external reviewer Anna McCord (ODI) for her helpful and relevant comments. We would also like to thank a number of other external reviewers for their comments: Laila Lokosang (African Union), Mariam Sow Soumare (NEPAD), Gaspar Fajth (UNICEF), Colin Andrews (World Bank), Martina Bergthaller (GIZ), Peter Ragno (UNICEF) and Maya Takagi (Secretaria de Relacoes Internacionais, of the government of Brazil). Special thanks go to Ruth Raymond for her editorial work.

Contents Acknowledgments Executive summary Abbreviations iii v ix 1. Introduction 1 1.1 Organization of the paper 5 2. Framework for linking social protection and agriculture 6 2.1 Introduction 6 2.2 Theoretical models 7 2.3 Theory of change 9 3. Methodology and sources of evidence 18 3.1 Criteria for inclusion 18 3.2 Outcomes of analysis 18 3.3 Results of the search 22 4. Impact of social protection on agriculture 26 4.1 Outcomes directly related to farm production 26 4.2 Outcomes indirectly related to farm production 33 4.3 Effects on the local economy 38 4.4 Impacts of social protection interventions on other economic outcomes 43 4.5 Gaps and weaknesses in the empirical literature 45 5. Impact of agriculture on risks and vulnerability 48 5.1 Outcomes related to the reduction of vulnerability and risks 48 5.2 Outcomes related to income generation capability 54 5.3 Effects on the local economy 58 5.4 Gaps and weaknesses in the empirical literature 60

6. Policy and research implications 62 6.1 Policy implications 62 6.2 Research implications 65 6.3 Conclusion 66 References 67 Annex 81

Executive Summary Introduction Social protection policies aim to reduce socio-economic risks, vulnerability, extreme poverty and deprivation, while smallholder agricultural policies focus on improving productivity in crops, fisheries, forestry and livestock and improving access to markets. Both areas of policy are important in poverty reduction strategies, but little attention has been paid to the interaction between them and how that influences their design and implementation. Conceptually, there is a two-way relationship between social protection and agriculture. On the one hand, poor rural households that mostly rely on agriculture for their livelihoods are often affected by limited access to resources, low agricultural productivity, poorly functioning markets and repeated exposure to covariate and idiosyncratic risks. Social protection can help to alleviate credit, savings and liquidity constraints by providing cash and in-kind support. In addition, the regularity and predictability of social protection instruments help households to better manage risks and to engage in more profitable livelihood and agricultural activities. On the other hand, agricultural policies and programmes can help smallholder households manage risk by stimulating farm output, income and overall household welfare. Since social protection and smallholder agricultural interventions often cover the same geographic areas and target the same households, there are opportunities for synergies and complementarities that would strengthen the livelihoods of poor rural households. This study explores the interaction between formal social protection and agriculture by proposing a theory of change and conducting an empirical review that identifies how social protection impacts agricultural production and how agricultural interventions reduce risks and vulnerability at the household and local economy levels. The paper seeks to provide an empirical rationale for building synergies and coordinating complementarities between social protection and smallholder agriculture in developing countries, especially in sub-saharan Africa. The review also provides some insights to the FAO and its partners on how social protection and agriculture can potentially complement each other. Methods This study reviews published and unpublished empirical literature and systematic reviews from the period 1999-2012. Evidence was collected from the impact evaluations of social protection and agricultural interventions in rural sub-saharan Africa, Latin America and Asia. The review is mainly based on the impact evaluations of non-contributory schemes, such as cash transfers, public works, school feeding, food aid, social pensions and education fee waivers. On the agricultural side, the review focused on land tenancy and titling, extension, irrigation, natural resource management, input technology, marketing arrangements, microfinance and rural infrastructure. The impact of social protection on agriculture Much of the evidence about the impact of social protection on agriculture comes from impact evaluations of cash transfers and public works schemes, although there is also some evidence from school feeding schemes and education fee waivers. v

Cash transfers and public works schemes can directly impact agriculture by increasing the accumulation of agricultural assets. However, the magnitude of accumulation varies based on programme design (value of transfer, regularity of payments, duration, complementary interventions), gender and socio-cultural context. While more evidence is required, available studies suggest that cash transfers and public works may increase expenditures in agricultural inputs, depending on the programme design (value of transfer, duration, predictability, initial endowment of wealth). A solid amount of evidence shows that social protection interventions reduce child labour and can make labour allocation decisions more flexible. Studies of cash transfers have reported shifts from on-farm to non-farm work in Latin America and shifts from casual work to ownfarm work in sub-saharan Africa. There are studies that have not found reductions in adult labour supply as a result of cash or in kind transfers. However, reductions in adult labour supply due to income effects have been reported for women, informal and unpaid workers who have received conditional cash transfers in Brazil and elderly beneficiaries of old age pensions in South Africa. Cash transfers and public works interventions also indirectly impact agriculture by preventing detrimental risk-coping strategies (e.g. selling ploughs or fishing equipment to buy food) and, together with school feeding and education fee waivers, by increasing investments in human capital development (e.g. child education and health services). The limited evidence available suggests that cash transfer interventions also increase off-farm investments in microenterprises. An emerging body of evidence shows that cash transfers and some public works interventions generate significant income multipliers in local economies as beneficiary households spend the transfers on goods and services mainly sold or produced by non-beneficiary households. However, complementary programmes for non-beneficiary households would be needed to relax capital and liquidity constraints and enable a better supply response that maximizes the income multiplier (e.g. microcredit). Some studies have found that cash transfers and old age pensions crowd out private transfers, while others have found the opposite and suggest that cash transfers schemes may improve participation in social networks. In the empirical literature, improvements in assets, inputs, agricultural output and off-farm investments are attributed to the alleviation of credit and liquidity constraints and the predictability of cash transfers and cash-based public works schemes. The alleviation of credit and liquidity constraints also encourages investments in human capital development, while the certainty of social protection mechanisms discourages adverse risk coping and makes labour allocation decisions more flexible. Consequently, changes in spending by the household generate additional income in the local economy. The impact of agriculture on risks, vulnerability and income generating capacity The evidence provides an empirical rationale as to how agriculture can complement social protection and points to potential synergies between the two sectors. Smallholder agricultural interventions can reduce household vulnerability and risk as measured by indicators of livelihood security. The evidence shows that many agricultural interventions increase household income and that interventions that improve access to microcredit, infrastructure, vi

irrigation, extension and input technology can lead to improvements in household consumption, food security and the accumulation of durable assets. Agricultural interventions can also improve the income generating capabilities of rural households, thus further protecting them from risks. Some studies have found that interventions that enhance certainty by guaranteeing access to land and/or insuring against potential crop losses have led to increased investments in high-return microenterprises. Several studies have suggested that irrigation and land tenancy reform increase the labour allocated to on-farm work. Microcredit schemes and some land tenancy reform measures increase the labour allocated to non-farm work. There is limited and inconclusive evidence concerning the impacts of agriculture on school attendance, risk coping and health. Cash grants for production and input subsidies may generate significant income multipliers throughout the local economy. In short, the nascent empirical literature appears to link various smallholder agricultural interventions to greater productivity and hence income, with subsequent increases in consumption and food security. Interventions that provide certainty and alleviate credit constraints may promote off-farm investments and investments in assets. In light of this emerging evidence, there is reason to believe that smallholder agricultural interventions can become social protection interventions in their own right if they are specifically targeted to the poorest and most vulnerable households. Knowledge gaps To better understand the role of social protection in agriculture, more evidence is needed concerning its impacts on crops, fishery, forestry and livestock production; the uptake of agricultural technologies to adapt to climate change; and natural resources management. There are numerous impact evaluations of conditional and unconditional cash transfer programmes in Latin America, but other regions have been less well studied. In addition, the interactions between multiple social protection interventions at local and national levels have generally not been addressed nor have the impacts of programmes that integrate social protection and agriculture. Similarly, most agricultural research does not follow the methodological standards needed for rigourous impact evaluation. Future studies of agricultural interventions should emphasize their impact on risk coping, informal risk management, human capital accumulation and the local economy. There is an important knowledge gap on the type of institutional arrangements at central and decentralized levels that could facilitate greater collaboration among the actors involved in social protection and agriculture. A comprehensive capacity development approach for stakeholders at the national level is needed to ensure greater coordination among social protection and agricultural policies and programmes. Conclusion Our review has shown that social protection and agriculture support the fulfilment of each other s objectives. We conclude that the available empirical literature suggests that there are potential synergies between social protection and agriculture at the household and local economy levels. That said, any efforts to capitalize on such synergies must be mindful of the conflicts that might arise from competition for power in the political economy of policy- vii

making. One should also be sensitive to programme design and implementation issues, such as inconsistent policy objectives, targeting, timing and coordination of interventions. Guidance on how to navigate these potential conflicts and maximize synergies would be beneficial to both policy-makers and practitioners. FAO s social protection agenda focuses on the interface between social protection, food and nutrition security, agriculture and livelihoods. Our paper is relevant to this agenda because it highlights the linkages between social protection and agriculture, which can be exploited and optimized toward building resilient and sustainable rural livelihoods. Our paper is also a timely contribution to the policy discourse on social protection and agricultural transformation in sub-saharan Africa. A major voice in this discourse is the CAADP 1 framework, which encourages synergies between social protection and agriculture that would enable agriculturedriven development to eradicate hunger and alleviate poverty and food insecurity. It is hoped that our review will be used as a starting point for future research and the formulation of guidance material. 1 Comprehensive Africa Agriculture Development Programme. The programme s goal is to eliminate hunger and reduce poverty through African governments allocating a minimum of 10 percent of their national budgets to agriculture and agreeing to raise annual agricultural productivity by a minimum of 6 percent (CAADP 2012). viii

Abbreviations AC Atención a Crisis (Nicaragua) BDH CCT BONOSOL CT DECT NREGS FA FFE FFW JPS OAP PAL PATH PETI PRAF PROCAMPO PROGRESA PSNP RPS UCT SAM CT-OVC LEAP PW Bono de Desarrollo Humano (Ecuador) Conditional cash transfer Bono Solidario (Bolivia) Cash Transfer Dowa Emergency Cash Transfer scheme (Malawi) National Rural Employment Guarantee Scheme (India) Familias en Accion (Colombia) Food for education Food for work Jaring Pengamanan Sosial (Indonesia) Old age pension Programa Apoyo Alimentario (Food Support Programme, (Mexico) Programme of Advancement through Health and Education (Jamaica) Programa de Erradicacao do Trabalho Infantil (Program to Erradicate Child Labor, Brazil) Programa de Asignacion Familiar (Family Allowance Programme, Honduras) Programa de Apoyos Directos al Campo (Programme for Direct Assistance in Agriculture, Mexico) Programa Nacional de Educacion, Salud y Alimentacion (National Programme for Education, Health and Nutrition, Mexico) Productive Safety Nets Programme (Ethiopia) Red de Proteccion Social (Social Protection Programme, Nicaragua) Unconditional cash transfer Social accounting matrix Cash Transfer Programme for Orphans and Vulnerable Children (Kenya) Livelihood Empowerment Against Poverty (Ghana) Public works/workfare programmes ix

1. Introduction Social protection is rapidly gaining prominence on the global development agenda, where it is being used to reduce vulnerability and food insecurity in the developing world. Recent and ongoing global crises (food prices, financial and debt crises and climate change) have heightened volatility and uncertainty in the global economy with far reaching consequences, especially for poor households in low-income countries (HLPE 2012). This partly explains the use of social protection as a policy response. In sub-saharan Africa and other regions, where households are continually susceptible to a wide variety of risks and shocks, social protection is becoming an important development tool. Since 2004, African Union (AU) member states have identified social protection as a key strategy for enhancing social development. AU member states have committed themselves to the Ouagadougou Declaration and Plan of Action to strengthen social protection schemes, increase their coverage and effectiveness for everyone, especially the poorest, most vulnerable and excluded persons (African Union 2008a). Further commitments include the 2006 Livingstone and Yaoundé Calls for Action on social protection, agreements at the 2007 International Labour Organization (ILO) regional meeting in Addis Ababa and recommendations of the 2008 regional meetings on Investing in Social Protection in Africa. All of these actions culminated with the development of a social policy framework for Africa (African Union 2008a). At the global level, some development agencies have begun to promote the development and strengthening of integrated social protection systems through multi-sectoral approaches (Rawlings et al. 2013). Yet one area that has received little attention in both policy discourse and empirical literature is the link between social protection and agriculture. This link is important and particularly relevant for rural households for several reasons. First, most rural beneficiaries of social protection are self-employed or rely on agriculture for their livelihoods. This means that, in practice, beneficiaries of agriculture and social protection policies may overlap and share the same geographical space. The extent of this overlap is debatable. Ostensibly, social protection would have little impact on the agricultural production of extremely poor, landless and labour-constrained beneficiaries. However, profiles of beneficiary households show that even the landless poor directly engage in agricultural activities by renting or borrowing land, sharecropping or providing agricultural wage labour (Davis and Debwre 2013; Asfaw et al. 2012b). While most agricultural interventions that target smallholders usually exclude the poorest households, they can still affect the productive behaviour of these households by bringing about changes in local agricultural labour markets. Nonetheless, agricultural interventions can potentially have a greater role in reducing vulnerability if they are targeted to the poorest households. Agricultural policy may have implications for the effectiveness of social protection policy and interventions, just as social protection policy may have implications for rural and agricultural livelihoods. Importantly, it is unlikely that social protection alone can sustainably lift households out of poverty, as it may not lead to long-term changes in livelihoods. Thus, there is a need to formulate social protection and complementary livelihoods promotion policies as part of a comprehensive rural development strategy. The well-documented role of agriculture in development and poverty reduction makes it a natural ally and complement to social protection. When combined, the two approaches can serve both immediate and long-term livelihood needs. 1

In most countries, agriculture has historically been the lead contributing sector in the early stages of growth and development (Diao et al. 2007). In sub-saharan Africa, agriculture constitutes one third of GDP, 50percent of total export value and remains the main source of livelihoods and food for two-thirds of the population. Approximately 72percent of the active rural population in the region are smallholder farmers and women comprise about half of the agricultural labour force (World Bank 2007). Agriculture makes key contributions to poverty reduction and development through various pathways. These include: increasing agricultural profits and labour income; increasing nonfarm rural labour incomes, non-farm profits and savings from investment in rural and urban areas (Hart 1998); lowering food prices, hence making tradable sectors more competitive; and tightening urban and rural labour markets, leading to higher unskilled wages in the economy (World Bank 2006). In China, evidence shows that GDP growth originating in agriculture was up to four times more effective in reducing poverty than GDP growth from non-agricultural sectors (Ravallion et al. 2007). The agricultural sector also plays a vital role in ensuring longterm food and nutrition security. Although agriculture is a key driver of development, there are several impediments to its growth and maturity in sub-saharan Africa. Agricultural growth has been slow and beset by challenges, including poor soil fertility and high variability in agro-ecological environments and farming systems; reliance on rain and traditional cultivation practices; underinvestment in human, physical and institutional capital; weak input and output markets; and unfavourable public policies, such as low spending, high taxation and structural adjustment (World Bank 2007; Diao et al. 2007). Recently, the revitalization of sub-saharan African agriculture has become a major focus of policy. One of several initiatives aimed at igniting rapid agricultural growth in the region is the Comprehensive Africa Agriculture Development Programme (CAADP) whose broad aim is to help African countries boost their economic growth and improve their food security through agriculture-led development. 2 A major challenge to agricultural growth and development in sub-saharan Africa is the fact that rural or agricultural households remain vulnerable to shocks and risks. Most rural households in low-income countries are net buyers of food, which makes them susceptible to production and market-related risks (de Janvry and Sadoulet 2011). Covariate risks, such as natural disasters, livestock diseases, climate change, financial crises, global food price hikes, conflict, economic collapse and devastating epidemics like HIV/AIDS, are major threats to the welfare of rural households (Dorward et al. 2006; Dercon 2005). In an uninsured rural population, exposure to idiosyncratic shocks such as illness, job loss, family deaths, births, migration, marriages and accidents can cause or deepen poverty. Rural households are often constrained by a lack of resources and low productivity, which makes it harder for them to cope with such risks and shocks. Furthermore, rural households often participate in imperfect and thin markets characterized by low economic activity and weak coordination (Dorward et al. 2006). Rural people are often poorly organized, which limits their access to markets and to adequate prices; while technical support services are weak or sometimes lacking. Consequently, the prevalent uncertainty, poverty and exposure to repeated shocks in rural 2 The programme s goal is to eliminate hunger and reduce poverty through African governments allocating a minimum of 10 percent of national budgets to agriculture and agreeing to raise annual agricultural productivity by a minimum of 6 percent (CAADP 2012). 2

areas all lead to a high degree of risk and vulnerability, hence the need for social protection (HLPE 2012; Dorward et al. 2006). Social protection interventions, when regular and predictable, are ideally suited to reducing the vulnerability of rural households by increasing consumption and relaxing credit, liquidity and insurance constraints. There are many definitions of social protection, most of which focus on risk management and the assistance of poor people (HLPE 2012). Generally defined, social protection is a set of interventions whose objective is to reduce social and economic risk and vulnerability and to alleviate extreme poverty and deprivation. Devereux and Sabates-Wheeler (2004) define social protection as the set of all initiatives, both formal and informal, that provide social assistance to extremely poor individuals and households; social services to groups who need special care or would otherwise be denied access to basic services; social insurance to protect people against the risks and consequences of livelihood shocks; and social equity to protect people against social risks such as discrimination and abuse. The European Community defines social protection as a specific set of actions to address the vulnerability of people s life through social insurance, offering protection against risk and adversity throughout life; through social assistance, offering payments and in kind transfers to support and enable the poor; and through inclusion efforts that enhance the capability of the marginalised to access social insurance and assistance (European Community 2010). The African Union defines social protection as encompassing a range of public actions carried out by the state and others that address risk, vulnerability, discrimination and chronic poverty. The right to social security in childhood, old age and at times of disability is expressed in a range of international Human Rights Declarations and treaties. Social security transfers in the form of, for example, pensions, child benefit and disability allowances are considered to be core elements of a comprehensive social protection system (African Union 2008). Social protection has four important roles (Devereux and Sabates-Wheeler 2004). Preventive instruments can stave off deprivation, mitigate the impact of an adverse shock and avoid adverse risk-coping strategies. Examples include regular and predictable cash transfers, the elimination of user fees and contributory social insurance/security (pensions, health insurance, maternity, disability, unemployment benefits, etc.). Protective instruments promote recovery from shocks and provide relief from economic and social deprivation, including the alleviation of chronic and extreme poverty and food insecurity. Examples include cash transfers, public employment schemes, feeding programmes and humanitarian relief. Promotive instruments enhance asset accumulation, human capital and income earning capacity among the poor and marginalized. They include conditional and unconditional cash transfers (CCTs, UCTs), asset building and livelihood development, elimination of user fees, school feeding programmes, second chance education, skills training and integrated early childhood development. Finally, transformative instruments address power imbalances that create or sustain economic inequality and social exclusion. Examples include workers rights, antidiscrimination policies and laws to protect inheritance rights. (Devereux and Sabates- Wheeler 2004). Much of the focus in empirical literature and policy debates has been on the human development outcomes of social protection interventions such as health and nutrition. There has been little discussion of the agricultural outcomes of social protection and, likewise, minimal attention has been paid to the potential role of agriculture in reducing vulnerability and averting risks. Identifying linkages between social protection and agriculture is ever more important in light of the growing risks to rural development in low income countries due to globalization, climate change, lagging investment in agriculture, imperfect agricultural inputs 3

and outputs, inadequate resources for social protection and chronic food insecurity (Farrington et al. 2004). Building synergies between social protection and agriculture could be integral to the implementation of strategies and policies for mitigating these risks and achieving sustainable and long-term rural development. This paper seeks to provide an empirical rationale for building such synergies and exploiting the complementarities between social protection and smallholder agriculture in developing countries, particularly in sub-saharan Africa. The paper builds upon earlier work commissioned by FAO to conceptualize the links between social protection and smallholder agriculture (Devereux et al. 2008a). In it, we explore the interaction between formal social protection and agriculture by developing a theory of change and conducting an empirical review that identifies how social protection impacts smallholder agricultural production and how agricultural interventions reduce risks and vulnerability at the rural household and local economy levels. On the one hand, we assess how social protection could produce impacts on agricultural production (crops, livestock, fishing and forestry). The evidence for this is primarily obtained from impact evaluations of social assistance instruments, such as cash transfer programmes, where possible augmented with findings from other instruments like public works, school feeding programmes, food aid, social pensions and education waivers. On the other hand, we determine how agricultural interventions can reduce risks and vulnerability and enhance income-generating capacity: impacts normally associated with social protection measures. Empirical evidence is obtained from the impact evaluations of community and smallholder-targeted agricultural interventions, which are classified into nine categories: land tenancy and titling; extension (including farmer field schools); irrigation; natural resource management; input technology (chemical, seed, implements, etc.); marketing arrangements (contract farming, cropping schemes, producer organizations); financial services (microfinance, crop insurance); transfers and subsidies (cash transfers for inputs, input fairs, input subsidies); and infrastructure. 3 We also use this evidence to determine whether there are potential synergies between social protection and agricultural interventions and consider the implications for policy and programme design. The study reviews evidence from developing regions, including sub-saharan Africa and Asia. We also include information from Latin America, where there is a large body of literature on the impact of social protection. The aim of this paper is to provide guidance for FAO s work on social protection and to contribute to the policy discourse on building resilience in rural communities and promoting sustainable rural livelihoods. Social protection is relevant to FAO s mandate to raise levels of nutrition, improve agricultural productivity, better the lives of rural populations and contribute to the growth of the world economy. FAO already promotes the role of social protection in enhancing food security and rural development. The twin-track approach adopted by FAO aims to i) improve livelihoods by promoting agricultural production and rural development with a focus on smallholders within a medium to long-term perspective, while also ii) facilitating more direct and immediate access to food, partly through social protection interventions (FAO 2003). As we will show, social protection can work on both tracks, since it can help to create the enabling environment for rural development. FAO s twin-track should thus be viewed as complementary approaches, which need to be carefully coordinated. FAO s comparative strengths in promoting agriculture, food security and rural 3 Adapted from IEG (2011b) 4

development make it the ideal agency to provide technical advice and build capacities to maximize the synergies between social protection and agriculture. 1.1 Organization of the paper The paper is organized in six sections. The first section is the Introduction. Section 2 presents the theoretical framework for linking social protection and agriculture. Section 3 presents the methods used in searching and selecting the appropriate empirical literature for review. Section 4 reviews the available evidence on the impact of social protection interventions on agriculture, while Section 5 reviews the available evidence on the impact of agricultural interventions on risks, vulnerability and income generating capacity. Section 6 concludes the paper by summarizing the evidence on the interaction between social protection and agriculture, highlighting the knowledge gaps and discussing the policy implications. 5

2. Framework for linking social protection and agriculture 2.1. Introduction Conventional economic theory has identified social protection as a public transfer characterized by income redistribution from the rich to the poor (Alderman and Yemtsov 2012). Accordingly, social protection has historically been viewed as having a negative effect on growth by reducing capital accumulation (via income redistribution), increasing fiscal deficits, producing a deadweight loss and economic distortions from the associated taxation and causing dependency and work and innovation disincentives (Alderman and Yemtsov 2012; Alesina and Perotti 1997). However, recent empirical evidence contradicts this theoretical orthodoxy, demonstrating that social protection interventions do not create work disincentives or dependency (Barrientos and Scott 2008; Abdulai et al. 2008). Moreover, the social protection goals of reducing vulnerability and managing risks are no longer viewed as interfering with growth objectives, but in fact as contributing to growth by increasing human capital accumulation and encouraging aggregate savings and risk-taking (Alderman and Yemtsov 2012; Ravallion 2007; Perotti 1993). There are good reasons to expect that social protection, and cash transfers in particular, can improve growth as a result of productive decisions at the household and local economy levels. The livelihoods of many beneficiaries of social protection programmes in sub-saharan Africa are predominantly based on subsistence agriculture and will continue to be so for the foreseeable future. The exit path from poverty is not necessarily the formal (or informal) labour market, but self-employment generated by beneficiary households themselves, whether within or outside of agriculture. Most social protection beneficiaries live in places where markets for financial services (such as credit and insurance), labour, goods and inputs are lacking, difficult to access and enter or do not function well. Cash transfers typically represent about 20 percent of their per capita expenditure and, when provided in a regular and predicable fashion, they may help households to overcome the obstacles that limit their access to credit or cash. This in turn can increase spending in productive and other incomegenerating activities, influence the role of the beneficiaries in social networks, increase their access to markets and inject resources into local economies. These impacts come through changes in household decisions and behaviour and, concomitantly, through impacts on the local economy of the communities (social networks, labour and goods markets, multiplier effects) where the transfers operate (Asfaw 2012). Agricultural interventions can also serve as instruments for social protection. Some agricultural interventions have an explicit social protection function since they are aimed at reducing risks, e.g. crop insurance, input subsidies, input grants and agricultural cash grants. At the micro level, agricultural interventions can improve agricultural output, household income, food security, risk coping, participation in social networks and a range of other welfare/social protection outcomes that reduce vulnerability and mitigate risks. At the meso and macro levels, agricultural growth lowers food prices and boosts food supplies, while the resulting profits increase the resources available for financing social protection. As mentioned previously, rapid agricultural growth produces an increase in rural labour incomes that may decrease the aggregate fiscal demand for social protection. 6

2.2 Theoretical models Several theoretical models can help predict how social protection leads to agricultural outcomes and how agriculture affects risks and vulnerability. While some of these models adhere to the traditional assumption that social protection instruments produce disincentives, they are still relevant because they provide a basis for hypothesizing pathways of impact. The agricultural household model is often used to describe the economic decision-making of rural households (Singh et al. 1986). In this model, when markets function perfectly production and consumption decisions can be viewed as separable. The model assumes that all prices are determined through market mechanisms and that households are price takers. Households suffer no labour, credit or other market constraints. There is no trade-off between the consumption of agricultural commodities and production for sale, since there are no transaction costs in food markets. In this context, agricultural households solve the profit maximization and utility maximization problem recursively by first maximizing profits from agricultural production based on standard economic theory, and second, given that profit, maximizing utility (Taylor and Adelman 2003). If the agricultural household model reflects reality, cash transfers should have little effect on agricultural production and instead only impact consumption (Boone et al. 2013). Yet, rural markets in developing countries do not often function perfectly. Liquidity and credit constraints are key factors leading poor agricultural households to lower than optimal use of types and quantities of inputs. Poor households, and particularly women, often find it difficult to borrow money due to a lack of assets to use as collateral or to credit rationing resulting from adverse selection, asymmetric information or government policies (Feder et al. 1990; Rosenzweig and Wolpin 1993; Fenwick and Lyne 1999; Lopez and Romano 2000; Barrett et al. 2001; Winter-Nelson and Temu 2005). Further, agricultural households in less developed regions often rely upon assets, such as livestock, as a form of savings or insurance, often an unsatisfactory risk-coping strategy for a variety of reasons. The lumpiness of these assets increases the difficulties of using them for savings and, in the context of covariate risk, they usually drop in value as many households try to sell them as a coping mechanism. Assets held as livestock or other animals are also risky due to the possibility of sickness or death. Rural markets are characterized by isolation or remoteness, with few buyers and sellers and covariate production risks (weather, prices, etc.). Moreover, high transaction costs in staple markets can often make self-sufficiency the best choice (Key et al. 2000). Transportation costs time and information-gathering activities add to the cost of selling food, creating a difference between the selling and buying prices; the more rudimentary the market, the higher these costs can be. Transaction costs, such as worker supervision, can lead households to use family workers over hired labour, although family and hired labour are imperfect substitutes (Deolalikar and Vijverberg 1987; Schmitt 1991). Further challenges to agricultural households include a lack of access to factors of production, including technology and extension services. Without access to adequate credit markets and with poor alternative risk-coping mechanisms, agricultural households often adopt low-risk low-return income generation strategies, often selling more than the optimal amount of labour off-farm in casual or exploitative labour markets in order to secure diverse sources of income (Dercon 2002). Due to multiple market failures, agricultural households may also make decisions based on ensuring that they have enough food to eat rather than on what would be the most profitable outcome (Boone et al. 7

forthcoming). For example, in the context of the risk of high food prices or thin product markets, households may prioritize the production of staple crops to ensure food security instead of producing more profitable cash crops. In the face of such constraints, the production and consumption decisions of agricultural households can be viewed as jointly determined, or non-separable (Singh et al. 1986). If household production and consumption decisions are non-separable, social protection and agricultural interventions may help overcome some of the constraints faced by agricultural households in the context of imperfect markets (Asfaw et al. 2012a). Cash transfers, for example, provide a guaranteed steady source of income if delivered at regular intervals, thus potentially making up for failures in the insurance market. This guarantee, especially for agricultural households that are less likely to have regular sources of income, may allow them to adopt riskier production strategies with a higher rate of return or to reduce the use of adverse risk-coping strategies. Furthermore, cash transfers provide liquidity and thus can be used for productive activities or as evidence that households can repay their debts. This may allow farmers to move closer to the optimal level of inputs when credit markets have failed. Such expenditures can be complemented by household labour and lead to increased agricultural or non-agricultural production. It is in this context that social protection instruments and agricultural interventions can help rural households alleviate some of the constraints and market failures that underlie non-separable consumption and production decisions. As suggested by the agricultural household model, the imperfect markets, credit constraints and volatile incomes caused by widespread risks (Fiszbein and Schady 2009) also constrain human capital investments (education and health) by rural households. There is extensive literature demonstrating that human capital accumulation raises economic growth and agricultural productivity (Jamison and Lau 1982; Pudasaini 1983; Jamison and Moock 1984; Azhar 1991). Social protection programmes like cash transfers can relieve constraints on human capital accumulation by meeting the costs and increasing the incentives associated with investing in human capital, thereby smoothing income fluctuations and reducing inequality within households and communities (Fiszbein and Schady 2009). In particular, conditional cash transfers (CCTs) address family underinvestments in human capital that are driven by imperfect information, incomplete altruism and short-sightedness (Fiszbein and Schady 2009). Another traditional theoretical model, Becker s Time and Household Production Theory (Becker 1965), also helps to explain the impact of social protection instruments and agricultural interventions on labour supply decisions. The model suggests that household time allocation decisions are based on a trade-off between the time allocated to utility-generating activities, like domestic production and/or leisure, and the time allocated to wage labour, which generates income (Becker 1965). When a household receives an income transfer, it may prefer utility-generating activities rather than wage labour, i.e. the income effect resulting from the transfer may create disincentives for paid work and incentives for domestic production or leisure (Parker and Skoufias 2000; Kanbur et al 1994). The income effect of a social transfer may vary by gender. For instance, while female recipients of a transfer may shift their labour supply from the labour market to domestic household care and work, male recipients may increase their leisure time. The income transfer may also lead to a substitution effect, with adults compensating for any reductions in child labour associated with increased school attendance, which is usually part of the conditionality for some cash transfers (Parker and Skoufias 2000). 8

The models we have discussed so far are often classified as unitary in nature, i.e. the household makes joint consumption and production decisions as a single unit or agent, and the members have the same preferences (Alderman et al. 1995). However, household members might not have identical preferences and the distribution of resources is often unequal, which makes the intrahousehold demographic composition and balance of power relevant for outcomes (Browning et al. 1994; Browning and Chiappori 1998). Empirical research has strengthened the case for the collective model (Thomas 1990; Chiappori 1992), while other studies have highlighted how preferences for production and consumption decisions vary by gender (Haddad et al. 1997; Handa and Davis 2006; Fiszbein and Schady 2009; Yoong et al. 2012). The collective model has implications for understanding how social protection instruments or agricultural interventions can influence household decision-making (Yoong et al. 2012). The extent of the bargaining power possessed by an individual often depends on their share of resources or earned income in the household (Yoong et al. 2012). Accordingly, the impact of agricultural interventions or social protection instruments on intrahousehold resource allocation especially to women could largely be influenced by who owns the means of production or by the cultural norms defining gender roles and power in the household. As such, many social protection programmes such as most CCTs in Latin America target women in order to increase their bargaining power and capitalize on their preferences to invest in children and food security. Yoong et al. (2012) however, note that social norms or lack of legal rights for women could hinder the success of gender-targeting in social protection programmes. In summary, our discussion of theoretical models makes the case for pathways or channels through which social protection instruments and agriculture affect the consumption and production decisions of rural households, especially in the presence of market failures and other constraints. Some models also recognize the potential gender differential in outcomes due to the intrahousehold decision-making processes in rural households. 2.3 Theory of change 2.3.1 Pathways of impact Following the rationale of the agricultural household model, the central assumption behind our theory of change is that consumption and production decisions are not separable for rural households living in a context of missing or incomplete markets. Multiple market failures and credit constraints may lead to suboptimal human capital investments, while the lack of knowledge about improved agricultural practices, inputs and factors of production hamper agricultural production. Social protection and agricultural interventions can play a vital role in easing these constraints for rural households. The underlying principles of the agricultural household model and the other models previewed earlier help us identify potential pathways through which social protection interventions can affect agricultural outcomes and agricultural interventions can function as social protection instruments. Our theory of change is also premised on the notion that the impacts of social protection and agricultural policies are not parallel but are interlinked, such that they contribute to each other s objectives of reducing risks and enhancing agricultural production. 9